# Balancing Safety and Helpfulness in Healthcare AI Assistants through Iterative Preference Alignment

arXiv: https://arxiv.org/abs/2512.04210 · HTML: https://mlsanity.com/paper/2512.04210/

## Key facts

- Cited by 1 papers in the map (data through 14 Oct 2026).
- Placed in Safety alignment and jailbreak attacks, one of 10 directions in LLM alignment / RLHF / preference optimisation.
- Builds on 25 of its 50 references that are in the map.
- First posted on arXiv 3 Dec 2025.

## Placed in

- LLM alignment / RLHF / preference optimisation / [Safety alignment and jailbreak attacks](https://mlsanity.com/alignment/safety-alignment-and-jailbreak-attacks/)

## Builds on (most cited first, 25 of 25)

- [Training language models to follow instructions with human feedback](https://mlsanity.com/paper/2203.02155/) (2022)
- [Proximal Policy Optimization Algorithms](https://mlsanity.com/paper/1707.06347/) (2017)
- [The Llama 3 Herd of Models](https://mlsanity.com/paper/2407.21783/) (2024)
- [DeepSeekMath: Pushing the Limits of Mathematical Reasoning in Open Language Models](https://mlsanity.com/paper/2402.03300/) (2024)
- [Direct Preference Optimization: Your Language Model is Secretly a Reward Model](https://mlsanity.com/paper/2305.18290/) (2023)
- [Self-Refine: Iterative Refinement with Self-Feedback](https://mlsanity.com/paper/2303.17651/) (2023)
- [Constitutional AI: Harmlessness from AI Feedback](https://mlsanity.com/paper/2212.08073/) (2022)
- [Large Language Models Encode Clinical Knowledge](https://mlsanity.com/paper/2212.13138/) (2022)
- [PubMedQA: A Dataset for Biomedical Research Question Answering](https://mlsanity.com/paper/1909.06146/) (2019)
- [CRITIC: Large Language Models Can Self-Correct with Tool-Interactive Critiquing](https://mlsanity.com/paper/2305.11738/) (2023)
- [MedMCQA : A Large-scale Multi-Subject Multi-Choice Dataset for Medical domain Question Answering](https://mlsanity.com/paper/2203.14371/) (2022)
- [HealthBench: Evaluating Large Language Models Towards Improved Human Health](https://mlsanity.com/paper/2505.08775/) (2025)
- [PKU-SafeRLHF: Towards Multi-Level Safety Alignment for LLMs with Human Preference](https://mlsanity.com/paper/2406.15513/) (2024)
- [OR-Bench: An Over-Refusal Benchmark for Large Language Models](https://mlsanity.com/paper/2405.20947/) (2024)
- [LLMs Accelerate Annotation for Medical Information Extraction](https://mlsanity.com/paper/2312.02296/) (2023)
- [MedAlign: A Clinician-Generated Dataset for Instruction Following with Electronic Medical Records](https://mlsanity.com/paper/2308.14089/) (2023)
- [Insights into Alignment: Evaluating DPO and its Variants Across Multiple Tasks](https://mlsanity.com/paper/2404.14723/) (2024)
- [CARES: Comprehensive Evaluation of Safety and Adversarial Robustness in Medical LLMs](https://mlsanity.com/paper/2505.11413/) (2025)
- [Fairness through Difference Awareness: Measuring Desired Group Discrimination in LLMs](https://mlsanity.com/paper/2502.01926/) (2025)
- [Refining Input Guardrails: Enhancing LLM-as-a-Judge Efficiency Through Chain-of-Thought Fine-Tuning and Alignment](https://mlsanity.com/paper/2501.13080/) (2025)
- [Towards Safe AI Clinicians: A Comprehensive Study on Large Language Model Jailbreaking in Healthcare](https://mlsanity.com/paper/2501.18632/) (2025)
- [The Necessity of AI Audit Standards Boards](https://mlsanity.com/paper/2404.13060/) (2024)
- [Trustworthy Medical Question Answering: An Evaluation-Centric Survey](https://mlsanity.com/paper/2506.03659/) (2025)
- [How Reliable AI Chatbots are for Disease Prediction from Patient Complaints?](https://mlsanity.com/paper/2405.13219/) (2024)
- [Self-Review Framework for Enhancing Instruction Following Capability of LLM](https://mlsanity.com/paper/2507.05598/) (2025)

## Built on by (most cited first, 1 of 1)

- [Robust Critics: Defending LLMs Against Multi-Turn Attacks](https://mlsanity.com/paper/2607.20472/) (2026)

## Abstract (arXiv)

Large Language Models (LLMs) are increasingly used in healthcare, yet ensuring their safety and trustworthiness remains a barrier to deployment. Conversational medical assistants must avoid unsafe compliance without over-refusing benign queries. We present an iterative post-deployment alignment framework that applies Kahneman-Tversky Optimization (KTO) and Direct Preference Optimization (DPO) to refine models against domain-specific safety signals. Using the CARES-18K benchmark for adversarial robustness, we evaluate four LLMs (Llama-3B/8B, Meditron-8B, Mistral-7B) across multiple cycles. Our results show up to 42% improvement in safety-related metrics for harmful query detection, alongside interesting trade-offs against erroneous refusals, thereby exposing architecture-dependent calibration biases. We also perform ablation studies to identify when self-evaluation is reliable and when external or finetuned judges are necessary to maximize performance gains. Our findings underscore the importance of adopting best practices that balance patient safety, user trust, and clinical utility in the design of conversational medical assistants.